PRISMA Flow diagram

Synthesis

Import dataset

Pre-specified outcomes

  1. Maternal Death
  2. Liver morbidity (hematoma, rupture, failure)
  3. Acute pulmonary edema
  4. Acute renal failure
  5. Dialysis
  6. Platelet transfusion
  7. Perinatal death

Maternal Death

Pairwise meta-analysis

                         RR            95%-CI %W(common)
Fonseca 2005         3.0000 [0.3202; 28.1042]       25.4
Fonseca 2019         0.2141 [0.0106;  4.3328]       14.1
Katz 2008            0.8750 [0.1280;  5.9809]       34.5
Ozer 2009                NA                          0.0
Van Runnard 2006     0.3548 [0.0156;  8.0730]       13.0
Vigil-De Gracia 1997 0.3333 [0.0146;  7.6344]       13.0

Number of studies: k = 5
Number of observations: o = 449
Number of events: e = 12

                        RR           95%-CI     z p-value
Common effect model 0.7700 [0.2492; 2.3799] -0.45  0.6499

Quantifying heterogeneity:
 tau^2 = 0 [0.0000; 7.4236]; tau = 0 [0.0000; 2.7246]
 I^2 = 0.0% [0.0%; 79.2%]; H = 1.00 [1.00; 2.19]

Test of heterogeneity:
    Q d.f. p-value
 2.64    4  0.6192

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2
- Q-Profile method for confidence interval of tau^2 and tau
- Continuity correction of 0.5 in studies with zero cell frequencies

Risk of bias

Subgroup analysis: Antepartum vs postpartum vs mixed

Subgroup analysis: Dexa vs Beta vs Prednisolone

Sensitivity analysis: Low RoB

Sensitivity analysis: Studies with zero events included

Funnel plot

Acute pulmonary edema

                 RR            95%-CI %W(common)
Fonseca 2005 3.0000 [0.3202; 28.1042]       24.2
Fonseca 2019 0.3494 [0.0146;  8.3368]       12.0
Katz 2008    0.3500 [0.0711;  1.7238]       47.6
Ozer 2009    1.0000 [0.0655; 15.2598]       16.3

Number of studies: k = 4
Number of observations: o = 381
Number of events: e = 14

                        RR           95%-CI     z p-value
Common effect model 0.6975 [0.2323; 2.0944] -0.64  0.5207

Quantifying heterogeneity:
 tau^2 = 0.1164 [0.0000; 13.2337]; tau = 0.3412 [0.0000; 3.6378]
 I^2 = 0.0% [0.0%; 84.7%]; H = 1.00 [1.00; 2.56]

Test of heterogeneity:
    Q d.f. p-value
 2.60    3  0.4572

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2
- Q-Profile method for confidence interval of tau^2 and tau
- Continuity correction of 0.5 in studies with zero cell frequencies

Acute renal failure

                 RR           95%-CI %W(common)
Fonseca 2005 0.7500 [0.2754; 2.0428]       26.7
Fonseca 2019 0.1892 [0.0239; 1.4998]        6.3
Katz 2008    0.6563 [0.3025; 1.4238]       44.8
Ozer 2009    0.6667 [0.1198; 3.7087]        9.1
Yalcin 1998  1.0000 [0.2390; 4.1844]       13.1

Number of studies: k = 5
Number of observations: o = 406
Number of events: e = 53

                        RR           95%-CI     z p-value
Common effect model 0.6658 [0.3965; 1.1179] -1.54  0.1239

Quantifying heterogeneity:
 tau^2 = 0 [0.0000; 2.4607]; tau = 0 [0.0000; 1.5687]
 I^2 = 0.0% [0.0%; 79.2%]; H = 1.00 [1.00; 2.19]

Test of heterogeneity:
    Q d.f. p-value
 1.78    4  0.7753

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2
- Q-Profile method for confidence interval of tau^2 and tau

Dialysis

Number of observations: o = 60
Number of events: e = 1

              RR            95%-CI    z p-value
Ozer 2009 3.0000 [0.1271; 70.7833] 0.68  0.4958

Details:
- Continuity correction of 0.5

Liver morbidity

                     RR           95%-CI %W(common)
Ozer 2009        0.3333 [0.0141; 7.8648]       45.4
Van Runnard 2006 0.1521 [0.0085; 2.7116]       54.6

Number of studies: k = 2
Number of observations: o = 91
Number of events: e = 4

                        RR           95%-CI     z p-value
Common effect model 0.2171 [0.0258; 1.8257] -1.41  0.1598

Quantifying heterogeneity:
 tau^2 = 0; tau = 0; I^2 = 0.0%; H = 1.00

Test of heterogeneity:
    Q d.f. p-value
 0.13    1  0.7191

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2
- Continuity correction of 0.5 in studies with zero cell frequencies

Platelet transfusion

                 RR           95%-CI %W(common)
Fonseca 2005 1.2000 [0.5574; 2.5832]       40.5
Fonseca 2019 0.8571 [0.4556; 1.6126]       59.5

Number of studies: k = 2
Number of observations: o = 219
Number of events: e = 49

                        RR           95%-CI     z p-value
Common effect model 0.9821 [0.6031; 1.5994] -0.07  0.9422

Quantifying heterogeneity:
 tau^2 = 0; tau = 0; I^2 = 0.0%; H = 1.00

Test of heterogeneity:
    Q d.f. p-value
 0.44    1  0.5069

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2

Perinatal death

                     RR           95%-CI %W(common)
Magann 1994      0.3611 [0.0432; 3.0169]       28.3
Van Runnard 2006 0.7969 [0.2103; 3.0197]       71.7

Number of studies: k = 2
Number of observations: o = 58
Number of events: e = 11

                        RR           95%-CI     z p-value
Common effect model 0.6372 [0.2062; 1.9693] -0.78  0.4337

Quantifying heterogeneity:
 tau^2 = 0; tau = 0; I^2 = 0.0%; H = 1.00

Test of heterogeneity:
    Q d.f. p-value
 0.38    1  0.5359

Details on meta-analytical method:
- Inverse variance method
- Restricted maximum-likelihood estimator for tau^2